Naval Research Lab Washington DC

1
Crystal Detector Element (CDE)Development Studies

• July 2002
• W. Neil Johnson
• Naval Research Laboratory
• for
• Eric Grove (NRL)
• Mai-Anh Byron (Swales)

2
CDE Development

• Fabrication processes for CDE have been studied
  at NRL and selected for EM fabrication.
• CDE wrapped Csl Crystal bonded with photodiode
  assy (PDA) at each end.

• Processes defined since PDR
• Bonding of PDA to CsI xtal
• CTE mismatch between PDA and xtal causes stress
  in bond
• Hard epoxies fail, silicones (without primer)
  dont adhere
• Need soft, flexible bond with strong adhesion
• Solution Silicone elastomer with primer. Bond
  laid up in mold that defines geometry.
• Wrapping of xtal
• Material is 3M VM2000 non-metallic, specular,
  reflective film
• High light yield, good optical properties
• Material is stiff, difficult to fold or wrap
• Solution Mold in mandrel at 120C for 2 hrs at
  final shape

Dual PIN photodiode with flex cable photodiode
assembly
3
Scope of Bonding Studies

• Assumptions
• The bond shall be made with DC93-500 silicone
  encapsulant, with DC92-023 primer applied to all
  bonding surfaces.
• The bond thickness shall be 0.7 ? 0.1 mm.
• The bond material shall not extend past the edges
  of the diode, lest it interfere with the bumper.
• The crystal surface shall be roughened to improve
  adhesion.
• Parameters studied
• Method of bond lay-up
• Selected one-stage mold with mask to define bond
  geometry
• Encapsulant is injected into bounded volume
• Mix ratio of DC93-500 encapsulant
• Conclusion vendor-recommended 101 is soft
  enough after cure
• Amount and method of DC92-023 primer to be
  applied
• Conclusion measured volume of primer is spread
  with spatula
• Surface treatment of diode and xtal (factory
  polish or roughened)
• Conclusion diode to remain polished, xtal to be
  roughened

4
Bonding Methods Studied

• Four bonding methods studied at NRL and Swales
• Spacer method (at NRL).
• Deliver measured amount of 93-500 between diode
  and xtal held fixed distance apart.
• OK, but susceptible to measurement errors.
  Rejected.
• Two-stage method (at NRL).
• Create pad of 93-500 on diode face, cure for one
  day, wet surface of 93-500 and apply to xtal.
• OK, but susceptible to incomplete curing.
  Rejected.
• Two-stage method (at Swales).
• Similar to NRL version, but uses more 93-500 in
  2nd stage and more precise volume.
• OK, but complex. Rejected.
• One-stage method (at Swales).
• Form bond within break-apart Teflon mold holding
  diode and xtal.
• Makes very strong and precise bonds. Selected.

5
Bonding Method Selected

• One-stage mold
• Principles of method
• Bond is formed within break-apart Teflon mold
  holding diode and xtal
• Mold precisely defines bond thickness, area, and
  location on face of xtal
• Mold is sealed against xtal face to prevent
  leakage of silicone encapsulant
• Encapsulant is injected into bond cavity and
  allowed to flow until any trapped bubbles are
  removed
• Bonding procedure is written and configured
• Process Specification for the Bonding of the PIN
  Photodiode Subassembly to the CsI Crystals",
  LAT-PS-00385-01

6
Bonding fixtures

• Exploded view of PDA and Mold subassembly
• Mold defines location of bond on xtal
  (referenced to xtal corner)
• Mask defines bond thickness (0.7 mm) and area
  (covers epoxy resin)
• Seal and O-rings prevent spread of encapsulant
  outside bond area

Mask
Seal
Photodiode Assembly
Mold
O-Rings
Photodiode Assembly
Mask
Xtal
Bond
7
Bonding fixtures

• Fixture and molds for single crystal
• Independent bonding molds are created for each
  crystal end
• Support sleeve prevents distortion of crystal and
  registers mold to corner of crystal

Support Fixture
CsI(Tl) Crystal
Mold and PDA Subassembly
Flex Cable soldered to Photodiode
8
Bonding fixtures

• Bonding work station
• 12 bonding stations, six on each side
• Encapsulant is injected with CDE in vertical
  position

Worktable
Injection syringe for DC93-500
Injection performed at this face. Cure for 24
hours before flipping 180 to bond opposite
face.
Crystal Bond Set-Up Assy.6 mounted on near
side6 mounted on far side
9
Bond Process

• Process at Swales Aerospace
• Prepare Surfaces
• Clean factory-polished diode bond surface
• Clean roughened crystal surface
• Allow cleaned surfaces to air dry
• Position Crystal into Fixture and Prime
• Position xtal into fixture
• Adjust pads to align xtal into proper position
• Prime xtal surface with DC92-023
• Position Photodiode into Mold and Prime
• Assemble diode into mold cavity
• Create desired bond area by assembling mask to
  mold assembly
• Prime diode surface with DC92-023

Inlet Outlet Holes
Figure 1 Diode in Mold, Top Side of Mold
Diode in Cavity
Mask in Cavity
Figure 2 Diode Mask in Mold, Back Side of Mold
10
Bond Process (cont.)

• Join Mold and Xtal fixture
• Assemble fixture (which now contains crystal) to
  mold assembly
• Preload crystal to mask/diode surface to create a
  tight seal
• Apply Adhesive
• Prepare adhesive mixture and de-gas
• Slowly pour mixed/degassed adhesive into syringe
• Inject adhesive into mold inlet hole until the
  bond cavity is filled
• After Injection
• Wipe clean adhesive from areas around inlet and
  outlet hole
• Cure for 24 hours until tack free
• Remove bonded crystal diode from mold after 24
  hours
• Allow bond to finish 7-day cure

11
Mechanical Test Sample Bonds (All Bonds
Displayed No Bubbles Desired Bond Profile (Area
Thickness Bond Maintained in Ceramic Frame )
Figure 13 EM Diode on Xtal, Side View
Figure 14 EM Diode on Xtal, Front View
Figure 15 EM Diode on Xtal, Back View
Figure 16 EM Diode on Xtal, Side View
Figure 16 EM Diode on Xtal, Side View
Figure 17 EM Diode on Xtal, Front View
Figure 18 EM Diode on Xtal, Back View
Figure 19 EM Diode on Plexiglas, Back View
12
Light Yield

• Light yield tests were performed on sample xtals
  3 x 3 x 19 cm
• Requirement on CDE
• gt6000 e/MeV in large PIN photodiode
• Tests were performed on sample xtals
• Xtal size 3 x 3 x 19 cm
• EM photodiode
• Tetratex wrap
• Samples exposed to radioactive source with known
  line energy
• Measured yield (after 50 thermal cycles)
• 6200 7000 e/MeV
• Expect yield of gt7500 e/MeV for EM dimensions
  with VM2000 wrap

13
Thermal Cycling

• Bonds need to be mechanically strong and maintain
  optical properties
• To survive handling loads
• To survive thermal cycling in test, shipping, and
  storage
• How do we know a bond has survived?
• A bond can be mechanically strong and show no
  visual evidence of separation, at the same time
  that it has optically failed!
• Bond process qualification plan must include
  readout of scintillation light.
• Visual inspection is inadequate and unacceptable
  as proof of bond quality.
• Thermal cycling
• Full qualification range 30C to 50C
• Gradient 20C per hour (if much steeper, xtals
  cant keep up)
• Soak 1 hour at 30C, 1 hour at 50C
• Irradiate xtal sample with 137Cs, monitor 662 keV
  centroid after series of thermal cycles

14
Thermal Cycling

• Optical performance of test samples under thermal
  cycling
• Test samples are 3 x 3 x 3 cm cubes with single
  EM photodiode
• Test samples typically decline 5 from their
  initial light yields and reach plateau
• Expect this to translate to 10 decline in CDEs
• Optical properties ofbonds survive
  thermalcycling
• Bonding processcreates thermallystable bonds

15
Thermal Cycling

• Optical performance of longer xtal samples
• Longer xtals samples are 3 x 3 x 19 cm with EM
  diodes on both ends
• Typically decline 10 from initial light yield
• Expect same performance from CDEs
• Again, bondingprocess createsthermally
  stablebonds
• Adhesion problemis solved

16
Early Thermal Cycling Tests

• Prior to creation of stable bonding process
• 16 sample bonds
• Created with several methods early in bonding
  study
• 13 survive cycling
• 3 have optical failure
• Optical failures
• 15 degradation in test sample is equivalent to
  50 degradation in CDE.
• We extracted samples for mechanical strength
  tests after optical failure

Optical failures
17
Mechanical Strength Tests

• Two types of destructive tests have been
  performed at NRL
• Tensile strength
• CETIM requirement 10 N (2.2 lbf)
• Shear strength
• CETIM requirement 0.12 N/mm2 (8 lbf 35 N for
  EM diode)
• gt35 samples tested
• Samples are pulled or sheared to failure in
  Dynamic Load Test Stand

Piston
Piston
Diode
Bond
18
Tensile Strength Test

• Swales one-stage bond
• After optical failure in thermal cycling
• Visual inspection showed no evidence for air
  gaps, delamination, separation, peeling, etc.
• Subjected to tensile strength test
• Bond failed at 280 N.
• Bond failed at diode face, not at xtal face.
• Bond strength is gt25x requirement, even after it
  has degraded from excellent to poor optical
  contact.

19
Shear Strength Test

• Swales one-stage bond
• Subjected to shear strength test
• Bond failed at 240 N.
• Bond failed at diode face, not at xtal face.
• gt10 shear tests have been performed
• Failure occurs typ. at gt300 N

20
Crystal Wrapping (LAT-PS-00795-01)

• VM2000 specular film
• Gives 20-30 more light than standard diffusive
  white wraps (e.g. Tyvek, Tetratex)
• Stable, rigid material will not wet xtal surface
  as Teflon-based wraps can (e.g. Tetratex)
• Non-metallic, so no grounding issues
• Rigid, so awkward to wrap
• Tried rolling, folding, spiral winding
• All created loose, awkward wraps
• Solution Hot molding
• Method
• Form VM2000 around aluminum mandrel in xtal form
  (with chamfers)
• Heat at 120 C for 2 hrs
• Material takes and holds sharp corners of mandrel
• Trivial to wrap around xtal, simple Kapton tape
  tacking
• No loss in light yield or mechanical stability
  from hot molding

21
Wrapping Studies Conclusions

• VM2000 Max. Molding Temperature as Specified by
  3M is 125ºC for 5 hrs.
• At this temperature or any other temperature
  below, material properties (mechanical, physical
  and optical) are NOT compromised as long as
  material is mechanically supported.
• Temperatures Studied For VM2000 Molding
• 50ºC, 75ºC, 100ºC, 120ºC, 150ºC
• Duration of Time for Molding Process at all
  Temperatures
• 1 hour, 2 hrs., 3 hrs.
• Conclusions
• the higher and the longer the temperature, the
  more defined the wrapping form
• 150ºC is too high for any duration of time tested
• 100ºC and 120ºC for 3 hours and 2 hours
  respectively is ideal
• Molding VM2000 in oven allows wrap to be
  preformed days/weeks before actual wrapping
  process on real crystal has to be performed.

22
Wrapping Mold Set-Up
23
Wrap Chamfer Close-Up